通过对Gnutella网络拓扑实测数据的分析,得到了Gnutella网络的累积节点度分布、聚类系数、平均路径长度和rich-club系数等拓扑特征.在拓扑特征分析的基础上,通过引入适应度、邻居数限制和节点聚类等机制,提出了面向非结构化P2P网络拓扑的FCL(Fitness model with tunable Clustering and neighbours Limitation)模型,并给出了建模算法.通过实验分析,证明了该模型的有效性.
With the IPv4 addresses exhausting and IPv6 emerging, the Peer-to-Peer (P2P) overlay is becoming increasingly heterogeneous and complex: pure IPv4, dual stack and pure IPv6 hosts coexist, and the connectivity limitation between IPv4 and IPv6 hosts requires the overlay protocols to be fit for this hybrid situation. This paper sets out to answer the question of how to construct multicast tree on top of IPv4-IPv6 hybrid network. Our solution is a New Greedy Algorithm (NGA) which eliminates the problem of joining failure in the hybrid network and keeps the efficiency of greedy algorithm in tree construction. Simulation results show that our algorithm has excellent performance, which is very close to the optimal in many cases.